Difference between revisions of "Test"

From Pcrewiki
Jump to navigationJump to search
 
(61 intermediate revisions by the same user not shown)
Line 8: Line 8:
  
 
test_1881 #slot
 
test_1881 #slot
 +
 +
 +
= 2014 =
 +
==November 2==
 +
Fixed the fiber for G797, now has LMS signal. HV is 1370.2 V.
 +
W1139, anode cable resistance is 4 ohm.3
 +
 +
==November 1==
 +
Things to be prepared or bought:
 +
 +
NIM->NEMAL adapters for Trigger Group, NEMAL extension cables with the same length.
 +
 +
Summing module (probably UVA 120/120A), and a VME/CAMAC discriminator (Ask Sergey to borrow one).
 +
 +
G900, HV channel problem
 +
 +
G57, anode signal is good, no dynode signal. Original HV is 1397.6 V. Dynode signal is back when the voltage is raised by 20 V, 50 V and 100 V. Finally fixed at 1447.6 V.
 +
 +
G386, anode signal is good, no dynode signal. Original HV is 1335.0 V. Dynode signal is back when the voltage is raised by 200 V and 250 V. Finally fixed at 1585.0 V.
 +
 +
G797, no anode signal, no dynode signal. Original HV is 1370.2 V. Cosmic appears when the voltage is raised by 200 V. Finally fixed at 1570.2 V, need to change the fiber.
 +
 +
W6, bad dynode signal, short tail, live with it.
 +
 +
W82, no dynode signal. Original HV is 967.0 V. Dynode signal is back when the voltage is raised by 20 V and 40 V. Finally fixed at 1007.0 V.
 +
 +
W160, bad dynode signal, long tail.
 +
 +
W243, bad dynode signal, long tail.
 +
 +
W347, cable resistance is 0, temporarily fixed, loose connection. HV 960.0 V -> 980.0 V
 +
 +
W979, no dynode signal. Original HV is 924.2 V. Dynode signal is back when the voltage is raised by 20 V and 40 V. Finally fixed at 964.2 V, long tail.
 +
 +
G245, cable resistance is infinite, fixed. Original HV is 1364.0 V, fixed at 1464.0 V.
 +
 +
W1150, no dynode signal. Original HV is 989.8 V. Dynode signal is back when the voltage is raised by 20 V and 40 V. Finally fixed at 1029.8, long tail.
 +
 +
==October 31==
 +
G357 is tested, and good now.
 +
G8, check the divider.
 +
W690, tolerable, check it during the calibration.
 +
G797, totally dead, try to change on Saturday (Nov. 1st).
 +
W1139, no anode signal but have dynode signal, check the divider and anode cable.
 +
 +
==August 04==
 +
1st version of readout list from Brad (hsft01.crl) is modified and successfully loaded.
 +
 +
Modules should work in Multi-Block Mode, so there should be more than 2 ADC boards in the crate.
 +
 +
A simple replay code is developed and available at cpeng@ifarm
  
  
 
==August 05==
 
==August 05==
−
Run Number: 91
+
Readout list is thoroughly tested. It can take data from up to 3 1881M ADC boards.
 +
 
 +
Some limit and problems are:
 +
 
 +
#More than 4 ADC boards would result in an "exception of interrupt" and reboot of the single board computer.
 +
#Need to update the crate map in the readout list (how many boards and at which slots they are)
 +
 
 +
Replay code is refined, it can automatically divide the data for each channel on each ADC board, right now it can only decode for a crate with up to 3 ADC cards.
 +
 
 +
'''Run Number: 91'''
  
 
ADC Number: 26, Slot 4.
 
ADC Number: 26, Slot 4.
Line 17: Line 77:
 
ADC Number: 28, Slot 6.
 
ADC Number: 28, Slot 6.
  
−
Run Number: 98
+
'''Run Number: 98'''
 +
 
 +
ADC Number: ??, Slot 9.
 +
ADC Number: ??, Slot 10.
 +
ADC Number: ??, Slot 11.
  
−
ADC Number: , Slot 9.
 
−
ADC Number: , Slot 10.
 
−
ADC Number: , Slot 11.
 
  
 
==August 06==
 
==August 06==
−
Slot 13 in the test crate (the most left) is found problematic
+
Readout list (pradtest.crl) is upgraded, it solves several problems, including:
−
ADC #30 is found problematic
+
#Now it supports up to 10 ADC boards for one crate, it is also the limit of our new power supply
 +
#Now it scan the whole crate and generate the crate map automatically, so no need to update the crate map in the code
 +
 
 +
The limit of this readout list is:
 +
 
 +
#It can only be used for ADC 1881M boards, no fastbus TDC is supported (which is not needed in our experiment)
 +
#There should be more than 2 ADC boards in the crate to work on the Multi-block Mode
 +
 
 +
The replay code is upgraded accordingly:
 +
#It supports up to 10 ADC boards, 1 tree for each. There are 65 branches (1 for total, 64 for each channel) in each tree
 +
#It now reads a configuration file at "./configuration/run_%d.dat" (or "./configuration/default.dat" if there is no such file), the configuration file tells the code how many ADC boards there are and the index number of the ADC boards
  
  
Line 57: Line 128:
 
board 11: There are uniform background with a width of 80 for each channel
 
board 11: There are uniform background with a width of 80 for each channel
  
−
'''Double check, run number: 108'''
+
'''Change slots and check again, run number: 108'''
  
 
Change slots and use the problematic boards only
 
Change slots and use the problematic boards only
Line 68: Line 139:
  
  
−
'''Run Number: 109'''
+
'''Run Number: 110'''
  
 
Slots: 6, 7, 8, 9, 10, 11, 12, 13, 14, 15
 
Slots: 6, 7, 8, 9, 10, 11, 12, 13, 14, 15
Line 74: Line 145:
 
ADC Number: 13, 14, 15, 16, 17, 18, 19, 20, 22, 23
 
ADC Number: 13, 14, 15, 16, 17, 18, 19, 20, 22, 23
  
−
$#$21 is TDC 1877, will not be used in this experiment
+
Number 21 is TDC 1877, will not be used in this experiment
  
 
All boards passed the test
 
All boards passed the test
 +
 +
Some events are bad on boards at slot 9, 11, 13. The installation is not very tight, it might be the reason.
 +
 +
'''Re-insert the boards Run Number: 111'''
 +
 +
Bad events are gone, but the channel 32 on both of board 18 and 20 has no readout. (The same as they are in Run 110)
 +
 +
18 and 20 are left to be tested again in the next run
 +
 +
 +
'''Run Number: 112'''
 +
 +
Slots: 6, 7, 8, 9, 10, 11, 12, 13, 14, 15
 +
 +
ADC Number: 18, 20, 24, 25, 26, 27, 28, 29, 30, 31 (additional board, numbered as 31, no mark on the board)
 +
 +
Board 18, channel 32 has no readout, it fails in 3 tests (run 110, 111, and 112) in a row.
 +
 +
Board 20, channel 32 has readout now, but all of its channels have a wide spread of pedestal. width is about 20 ~ 25, has a slope at the left.
 +
 +
 +
 +
==August 11==
 +
 +
The left crate is tested again, there is still a problem with recognizing.
 +
 +
Right crate is setup as primexroc6, center crate is setup as primexroc5, left crate will be setup as primexroc4
 +
 +
Test with 250 ns ADC gate: '''Run Number 121''', '''Run Number 122'''
 +
 +
According to '''Run Number 122''', center crate has a noise problem, try to fix it.
 +
 +
Noise problem fixed by cleanup the redundant cabling. The problem with #18 and #20 is found to be a mistake on decoding, actually the problem is caused by #31, which is an additional board with out primex mark.
 +
 +
In today's test:
 +
 +
# ADC_12, channel 58 no readout
 +
# ADC_15, channel 32 no readout, but it passes the test with readout on channel 32
 +
# ADC_17, channel 32 no readout, but it passes the test with readout on channel 32
 +
 +
'''Run Number 126''' Replace these 3 boards with #10, #9, #8, problems of no readout are gone, these 3 boards may need to be replaced.
 +
 +
'''Run Number 127''' ADC_8 is found quite noisy, replace with ADC_7. All good now. Will test ADC_8 in future.
 +
 +
 +
==August 13==
 +
 +
All crates are working now.
 +
 +
=2016=
 +
==January 15=
 +
LMS check, 607
 +
Cosmic run, 608

Latest revision as of 11:35, 15 January 2016

test fastbus crates by PPC

ld </site/coda/2.6.2/VXWORKSPPC/bin/fb_diag_cl.o

fb_diag

ld </site/coda/2.6.2/VXWORKSPPC/bin/test_1881.o

test_1881 #slot


2014

November 2

Fixed the fiber for G797, now has LMS signal. HV is 1370.2 V. W1139, anode cable resistance is 4 ohm.3

November 1

Things to be prepared or bought:

NIM->NEMAL adapters for Trigger Group, NEMAL extension cables with the same length.

Summing module (probably UVA 120/120A), and a VME/CAMAC discriminator (Ask Sergey to borrow one).

G900, HV channel problem

G57, anode signal is good, no dynode signal. Original HV is 1397.6 V. Dynode signal is back when the voltage is raised by 20 V, 50 V and 100 V. Finally fixed at 1447.6 V.

G386, anode signal is good, no dynode signal. Original HV is 1335.0 V. Dynode signal is back when the voltage is raised by 200 V and 250 V. Finally fixed at 1585.0 V.

G797, no anode signal, no dynode signal. Original HV is 1370.2 V. Cosmic appears when the voltage is raised by 200 V. Finally fixed at 1570.2 V, need to change the fiber.

W6, bad dynode signal, short tail, live with it.

W82, no dynode signal. Original HV is 967.0 V. Dynode signal is back when the voltage is raised by 20 V and 40 V. Finally fixed at 1007.0 V.

W160, bad dynode signal, long tail.

W243, bad dynode signal, long tail.

W347, cable resistance is 0, temporarily fixed, loose connection. HV 960.0 V -> 980.0 V

W979, no dynode signal. Original HV is 924.2 V. Dynode signal is back when the voltage is raised by 20 V and 40 V. Finally fixed at 964.2 V, long tail.

G245, cable resistance is infinite, fixed. Original HV is 1364.0 V, fixed at 1464.0 V.

W1150, no dynode signal. Original HV is 989.8 V. Dynode signal is back when the voltage is raised by 20 V and 40 V. Finally fixed at 1029.8, long tail.

October 31

G357 is tested, and good now. G8, check the divider. W690, tolerable, check it during the calibration. G797, totally dead, try to change on Saturday (Nov. 1st). W1139, no anode signal but have dynode signal, check the divider and anode cable.

August 04

1st version of readout list from Brad (hsft01.crl) is modified and successfully loaded.

Modules should work in Multi-Block Mode, so there should be more than 2 ADC boards in the crate.

A simple replay code is developed and available at cpeng@ifarm


August 05

Readout list is thoroughly tested. It can take data from up to 3 1881M ADC boards.

Some limit and problems are:

  1. More than 4 ADC boards would result in an "exception of interrupt" and reboot of the single board computer.
  2. Need to update the crate map in the readout list (how many boards and at which slots they are)

Replay code is refined, it can automatically divide the data for each channel on each ADC board, right now it can only decode for a crate with up to 3 ADC cards.

Run Number: 91

ADC Number: 26, Slot 4. ADC Number: 7, Slot 5. ADC Number: 28, Slot 6.

Run Number: 98

ADC Number: ??, Slot 9. ADC Number: ??, Slot 10. ADC Number: ??, Slot 11.


August 06

Readout list (pradtest.crl) is upgraded, it solves several problems, including:

  1. Now it supports up to 10 ADC boards for one crate, it is also the limit of our new power supply
  2. Now it scan the whole crate and generate the crate map automatically, so no need to update the crate map in the code

The limit of this readout list is:

  1. It can only be used for ADC 1881M boards, no fastbus TDC is supported (which is not needed in our experiment)
  2. There should be more than 2 ADC boards in the crate to work on the Multi-block Mode

The replay code is upgraded accordingly:

  1. It supports up to 10 ADC boards, 1 tree for each. There are 65 branches (1 for total, 64 for each channel) in each tree
  2. It now reads a configuration file at "./configuration/run_%d.dat" (or "./configuration/default.dat" if there is no such file), the configuration file tells the code how many ADC boards there are and the index number of the ADC boards


August 07

Test for all boards


Run Number: 106

Slots: 6, 7, 8, 9, 10, 11

ADC Number: 1, 2, 3, 4, 5, 6

All boards passed the test


Run Number: 107

Slots: 6, 7, 8, 9, 10, 11

ADC Number: 7, 8, 9, 10, 11, 12

All boards passed the test

board 8 was marked as "wide pedestal". No difference seen.

board 7: channel 58, no readout.

board 10: channels 16 ~ 31, 48 ~ 63, no readout

board 11: There are uniform background with a width of 80 for each channel

Change slots and check again, run number: 108

Change slots and use the problematic boards only

Slots: 14, 15, 16

ADC Number: 11, 7, 10

Every problem disappears


Run Number: 110

Slots: 6, 7, 8, 9, 10, 11, 12, 13, 14, 15

ADC Number: 13, 14, 15, 16, 17, 18, 19, 20, 22, 23

Number 21 is TDC 1877, will not be used in this experiment

All boards passed the test

Some events are bad on boards at slot 9, 11, 13. The installation is not very tight, it might be the reason.

Re-insert the boards Run Number: 111

Bad events are gone, but the channel 32 on both of board 18 and 20 has no readout. (The same as they are in Run 110)

18 and 20 are left to be tested again in the next run


Run Number: 112

Slots: 6, 7, 8, 9, 10, 11, 12, 13, 14, 15

ADC Number: 18, 20, 24, 25, 26, 27, 28, 29, 30, 31 (additional board, numbered as 31, no mark on the board)

Board 18, channel 32 has no readout, it fails in 3 tests (run 110, 111, and 112) in a row.

Board 20, channel 32 has readout now, but all of its channels have a wide spread of pedestal. width is about 20 ~ 25, has a slope at the left.


August 11

The left crate is tested again, there is still a problem with recognizing.

Right crate is setup as primexroc6, center crate is setup as primexroc5, left crate will be setup as primexroc4

Test with 250 ns ADC gate: Run Number 121, Run Number 122

According to Run Number 122, center crate has a noise problem, try to fix it.

Noise problem fixed by cleanup the redundant cabling. The problem with #18 and #20 is found to be a mistake on decoding, actually the problem is caused by #31, which is an additional board with out primex mark.

In today's test:

  1. ADC_12, channel 58 no readout
  2. ADC_15, channel 32 no readout, but it passes the test with readout on channel 32
  3. ADC_17, channel 32 no readout, but it passes the test with readout on channel 32

Run Number 126 Replace these 3 boards with #10, #9, #8, problems of no readout are gone, these 3 boards may need to be replaced.

Run Number 127 ADC_8 is found quite noisy, replace with ADC_7. All good now. Will test ADC_8 in future.


August 13

All crates are working now.

2016

=January 15

LMS check, 607 Cosmic run, 608

Retrieved from "https://wiki.jlab.org/pcrewiki/index.php?title=Test&oldid=936"